The Top Floor Flat

In October, 2012, I joined the team at 23snaps, a mobile app developer with a product that allows parents to save photos and videos of their children to a beautiful digital journal, and share those photos privately with family and close friends.

As head of marketing, one of my primary objectives was to acquire new users. I quickly discovered that the tools and networks for advertising mobile applications were very much in their infancy – so much so that the general feeling among mobile marketers was that there is no truly effective place for mobile ad spend, particular if you’re not promoting a mobile game*.

Any spend put towards generic mobile ad networks, which offer little or no targeting, is more likely to be written off as a failed experiment than to generate new engaged users, and targeted networks that allow demographic- or interest-based targeting are prohibitively expensive for early-stage developers, with ad agencies often requiring a minimum spend of $10,000 for a trial campaign.

Two early tests we ran at 23snaps on two different ad networks, one general and one more targeted, showed such poor results that we cancelled the campaigns early. Worryingly, with both campaigns, we saw generally high click rates, depleting our budget, but no new registrations. Whether this was due to the network incentivizing clicks from an irrelevant audience or fat fingers on a small screen, the end result was money out and no new users in.

At about the same time, Facebook, whose ad network has previously been criticized for being ineffective and a waste of money, announced a new advertising format. Facebook would allow mobile app developers to buy advertising space within the Facebook mobile application; with custom ads that delivered users directly into the Apple App Store or Google Play Store page for the advertiser’s app.

These ads would appear in a reader’s news feed, could be targeted using all of the interest, demographic and location-based targeting Facebook provides for any other ad format (including targeting ads based on pages and products a user has ‘Liked’ on Facebook already), and would be formatted to include an “Install now” button, information about how many of the reader’s friends user the app and a star rating. Despite our less than stellar results from other ad networks, we were intrigued, primarily because of the opportunity to deliver highly targeted ads to a relevant audience of our potential users.

We transferred a portion of our Facebook advertising budget to test this new format… and were absolutely astounded by the results. Not only were we seeing the number of app downloads jump significantly, the ads delivered the highest click through rate of any Facebook ad campaign we’d ever run, with conversion rates from click to download rivaling those of even our own website. For both Apple and Android devices, the Facebook ads were performing so much better than any third party advertising channel we had previously tested, or even heard of, that we increased the spend and haven’t looked back.

But Facebook isn’t the only social network that is working magic for mobile advertisers. YPlan, a London-based mobile app that allows Londoners to discover great events going on nearby, spent most of December running promoted Tweets on Twitter – that is paying Twitter to feature their message in the feed of relevant users, targeted based on location, topics of interest, who they follow and demographics. The result? YPlan popped up in the top 10 of all Lifestyle apps in the UK app store, a feat that usually requires anywhere between 1,000 and 5,000 downloads per day. When the Twitter ads stopped? YPlan dropped out of the top 100 Lifestyle apps.

Encouraged by the clear superiority of social networks in effectively delivering mobile ads to a relevant audience at a lower cost, I looked into one more company that, while not purely a social network, has intimate access to user’s social graphs and interests: Google. Like Facebook, Google’s AdWords advertising network has made some changes to offer better options for mobile advertisers. Not only can advertisers target ads based on app interest category (such as lifestyle, business, games or education), in Google’s case, advertisers can select individual applications (either competitors or apps that have overlapping user groups) in which to promote their own app.

While not all developers allow Google to serve ads in their applications, I was surprised by the high number that did; and discovered that for every ten or so applications I wanted to target, at least one would allow me to show my ads. Given there are millions of apps on the market, this still offers a pretty large range of advertising real estate.

Like Facebook, this highly targeted channel resulted in exceptionally cost-effective advertising. More importantly, Google (and Facebook) provided transparency into which ads were working, the cost per download associated with each of my campaigns and gave me the flexibility to change, or stop, my campaign at any time if it wasn’t meeting my cost per download targets.

This has all led me to believe social media networks (and companies that have access to the social graph online) are going to replace mobile ad networks as the most effective way to promote mobile applications. Social networks have access to incredibly detailed information about their users’ interests, demographics and background, allowing them to give advertisers much more effective targeting tools. Many users access these social network on their phones, allowing the networks to deliver ads where users can immediately download advertiser’s applications. They also have a readymade audience that outnumbers even the largest advertising network, which relies on finding publishers (i.e. other mobile developers) to offer sell ad inventory. Finally many of these social networks already have built in advertising platforms, making it cost-effective for them to allow smaller advertisers to create small, self-managed campaigns, as opposed to many ad agencies and large advertising networks that do not have the technology to allow advertisers to manage their own campaigns so much charge a premium to ensure they cover costs.

I’ll be incredibly interested to see how social networks continue to develop their mobile advertising offerings, and whether they continue to outperform ad networks as the industry grows something that my experience so far has indicated is exceptionally likely.

*I single out mobile games as there are some networks that claim to offer highly cost-effective advertising campaigns for mobile gaming apps. These ad networks work by allowing one game manufacturer to earn money by offering incentives in the form of in-game rewards to users who download other advertisers’ games. Not working with a game app myself, I can’t speak to this model or the actual cost of the service.